Silica-Based and Borate-Based, Titania

Journal of Functional Biomaterials

Article Silica-Based and Borate-Based, Titania-Containing Bioactive Coatings Characterization: Critical Strain Energy Release Rate, Residual Stresses, Hardness, and Thermal Expansion

Omar Rodriguez 1,2,*, Ali Matinmanesh 1,2, Sunjeev Phull 1, Emil H. Schemitsch 2,3, Paul Zalzal 4,5, Owen M. Clarkin 6, Marcello Papini 1 and Mark R. Towler 1,2,7 1 Department of Mechanical & Industrial Engineering, Ryerson University, Toronto, ON M5B 2K3, Canada; [email protected] (A.M.); [email protected] (S.P.); [email protected] (M.P.); [email protected] (M.R.T.) 2 St. Michael’s Hospital, Toronto, ON M5B 1W8, Canada; [email protected] 3 Department of Surgery, University of Western Ontario, London, ON N6A 4V2, Canada 4 Oakville Trafalgar Memorial Hospital, Oakville, ON L6J 3L7, Canada; [email protected] 5 Faculty of Health Sciences, Department of Surgery, McMaster University, Hamilton, ON L8S 4L8, Canada 6 School of Mechanical and Manufacturing Engineering, Dublin City University, Dublin D09 W6Y4, Ireland; [email protected] 7 Department of Biomedical Engineering, University of Malaya, Kuala Lumpur 50603, Malaysia * Correspondence: [email protected]; Tel.: +1-416-979-5000 (ext. x4518)

Academic Editor: Adriana Bigi Received: 12 October 2016; Accepted: 28 November 2016; Published: 1 December 2016

Abstract: Silica-based and borate-based glass series, with increasing amounts of TiO2 incorporated, are characterized in terms of their mechanical properties relevant to their use as metallic coating materials. It is observed that borate-based glasses exhibit CTE (Coefﬁcient of Thermal Expansion) closer to the substrate’s (Ti6Al4V) CTE, translating into higher mode I critical strain energy release rates of glasses and compressive residual stresses and strains at the coating/substrate interface, outperforming the silica-based glasses counterparts. An increase in the content of TiO2 in the glasses results in an increase in the mode I critical strain energy release rate for both the bulk glass and for the coating/substrate system, proving that the addition of TiO2 to the glass structure enhances its toughness, while decreasing its bulk hardness. Borate-based glass BRT3, with 15 mol % TiO2 incorporated, exhibits superior properties overall compared to the other proposed glasses in this work, as well as 45S5 Bioglass® and Pyrex.

Keywords: enameling; coefﬁcient of thermal expansion; borate-based glass; indentation

1. Introduction Direct skeletal attachment (DSA) is a method used in prosthetics in which a metallic implant is attached directly to the patient’s bone at the residual limb; concerns regarding DSA include infection and skin irritation [1–3]. Different approaches have been taken towards re-designing DSA devices for improving patient outcomes; these approaches usually involve modification of the surface by sandblasting the device surface, titanium plasma-spraying, plasma-spraying with hydroxyapatite (HA), coating the implant with a titanium dioxide (TiO2) layer through anodic oxidation, and applying a coating made from bioactive glass [4–6]. From these approaches, bioactive glasses have shown encouraging results over these other technologies when used as coatings [5]. The potential of bioactive glasses as coatings was first postulated with the development of Hench’s 45S5 Bioglass® in the 1960s [7]. Bioglass® was the first synthetic material to chemically

J. Funct. Biomater. 2016, 7, 32; doi:10.3390/jfb7040032 www.mdpi.com/journal/jfb J. Funct. Biomater. 2016, 7, 32 2 of 14 adhere to both hard and soft tissue [7]. Although bioactive glasses have been employed for coating metals [8–11], these compositions have all, to date, contained alumina [8,11], which is associated with both defective bone mineralization and neurotoxicity [12]. Other compositions have been deficient in zinc [9–11], an antibacterial component [13–15] which aids the healing process by inhibiting the growth of caries-related bacteria such as Streptococcus mutans [16]. This work considers two distinct glass series, one based on silica (SiO2), one based on borate (B2O3), with increasing amounts of titanium dioxide (TiO2) incorporated at the expense of silica and borate, respectively. B2O3 has been shown to reduce the coefficient of thermal expansion (CTE) of glasses [17], so that borate glasses have CTEs closer to that of the metallic substrate to be coated (typically Ti6Al4V, with a CTE of 9.5 × 10−6/◦C in the range of 0–315 ◦C[18]). Processing such glasses for use as coatings (e.g., through enameling [9], plasma spraying [19], electrophoretic deposition [20,21], or glazing [22]) requires heat treatment to allow for the glass to react with the substrate surface thus creating a chemical bond [23,24]. Once the bond has formed and the assembly is cooled, a difference in CTE between the glass and metal will induce residual stresses, hence causing cracks to appear in the glass or at the glass/substrate interface. For this reason, a borate-based glass series is proposed, to evaluate the effect of B2O3 on its coating capability by means of its reduced CTE compared to silica-based glasses; a silica-based glass series is also proposed, with a homologous composition to that of the borate-based glass series, to allow for the evaluation of the effect of B2O3 versus SiO2 on the resultant properties of the coating. Additionally, TiO2 is incorporated in these glasses as it helps promote a more stable chemical bond when coating such a glass onto Ti6Al4V [17]. Additionally, titanium is known to create a permanent bond to bone, via osseointegration [25,26]. In terms of metal coating techniques, enameling is probably one of the most widely used. Among the different findings in the literature, researchers have found that low firing temperatures and short sintering times did not help the glass to spread along the substrate surface, ultimately resulting in a very porous coating [8], that silica-based glasses with significantly high percentage of SiO2 exhibit better adhesion to the metallic substrate due to lesser thermal expansion mismatch between the glass and the substrate [27], and that smaller processing windows (the range between glass transition and crystallization temperature where the coating is heat treated to) favors crystallization hence reducing bioactivity of the glass [28]. Other coating methods have been explored, including reactive plasma spraying [19], electrophoretic deposition [20], and dip coating [29]. Good adhesion has been reported using these methods; however, these methods may require sintering at high temperatures [19,20,29], which may hinder the bioactive performance of glass due to crystallization, and they produce fragile coatings [29], compromising the mechanical stability of the coating. Indentation-based measurement methods allow a quick and qualitative measurement of the adhesion [11,30–32]. Such tests can also be used to quantify fracture toughness of the material by the direct measurements of cracking after indentation [33–39]. A common example is the Vickers indentation fracture (VIF) test, which measures the lengths of the cracks emanating from the Vickers indents. This technique was first developed by Lawn et al. [33], under the assumption that such cracks were created due to tensile stresses that form during unloading. Anstis et al. [34] validated Lawn’s model for several ceramics and glasses by comparing the fracture toughness obtained from the VIF test with the ones obtained from standard fracture tests. Later, Laugier [39] showed that indentation crack geometry in glasses and ceramics were different and claimed that Lawn’s model required some modifications when used for the evaluation of ceramic toughness, and therefore developed a new model that described the indentation cracking in ceramics more realistically. In this study, both glass series will be evaluated to determine their coefficient of thermal expansion (CTE) and its effect on the residual stresses and strains post-coating, their critical strain energy release rate in mode I (opening) of the coating/substrate system through double-cantilever beam (DCB) specimens and of the bulk glass through Vickers indentation, and their bulk hardness. J. Funct. Biomater. 2016, 7, 32 3 of 14

2. Results

2.1. Coefﬁcient of Thermal Expansion (CTE) J. Funct. Biomater. 2016, 7, 32 3 of 14 Results from the measurement of the CTE are found in Figure1 for the SRT (silica-based) and BRT (borate-based)J. Funct.Results Biomater. glass fr 2016om series. , the7, 32 measurement The CTE of of the the silica-based CTE are found glasses in Figure were 1 for found the SRT to be(silica consistently-based)3 ofand 14 greater BRT (borate-based) glass series. The CTE of the silica-based glasses were found to be consistently than the CTE of Ti6Al4V, with the percentage difference ranging between 11.1% and 24%, whereas the greaterResults than frtheom CTE the measurementof Ti6Al4V, with of the the CTE percentage are found differen in Figurece ranging 1 for the between SRT (silica 11.1%-based) and 24%, and CTE for the borate-based glasses were found to be below the CTE of Ti6Al4V with a smaller percentage BRTwhereas (borate the- based)CTE for glass the borateseries.- basedThe CTE glasses of the were silica found-based to glassesbe below were the found CTE ofto Ti6Al4Vbe consistently with a difference,greatersmaller ranging thanpercentage the between CTE difference, of 4.1%Ti6Al4V, andranging with 5.8%. betweenthe Inpercentage statistical 4.1% and differen terms,5.8%.ce In ranging the statistical CTE between for terms, all 11.1 borate-basedthe% CTE and for 24%, all glasses were foundwhereasborate to-based be the equivalent; CTEglasses for thewere theborate fo CTEund-based to for be SRT0glasses equivalent; and were SRT1 foundthe (0CTE to and befor 5below SRT0 mol% theand incorporated CTE SRT1 of Ti6Al4V(0 and TiO 5 withmol%2) were a also found tosmallerincorporated be equivalent. percentage TiO2) weredifference, also found ranging to be between equivalent. 4.1% and 5.8%. In statistical terms, the CTE for all borate-based glasses were found to be equivalent; the CTE for SRT0 and SRT1 (0 and 5 mol% incorporated TiO2) were also found to be equivalent.

Figure 1. CTE for the SRT and BRT glasses, plotted along with the CTE of Ti6Al4V as a reference. Figure 1. CTE for the SRT and BRT glasses, plotted along with the CTE of Ti6Al4Vas a reference. Scatter barsScatter indicate bars indicate one standardone standard deviation deviation from from thethe mean. Figure 1. CTE for the SRT and BRT glasses, plotted along with the CTE of Ti6Al4V as a reference. 2.2. Residual Stress and Strain Analysis 2.2. ResidualScatter Stress bars and indicate Strain one Analysis standard deviation from the mean. Residual strain results (Equation (3)) are shown in Figure 2, where it can be seen that, in terms Residual2.2of the. Residual magnitude, strain Stress results andgreater Strain (Equation strains Analysis were (3)) experienced are shown when in Figure the silica2, where-based coatings it can be were seen employed, that, in terms of the magnitude,opposedResidual to borate greater strain-based results strains ones. (Eq wereuationThis isexperienced (due3)) areto highershown CTEwhen in Figure mismatch the 2, silica-based where between it can the be coatings silica seen- that,based were in glasses terms employed, opposedofand tothe the borate-based magnitude, titanium substrat greater ones. e,strains Thisespecially iswere due forexperienced to the higher case of when CTE SRT3 the mismatch (15 silica mol- based% betweenincorporated coatings the were TiO silica-based2 ).employed, Residual glasses opposedstress in the to borateglass coating-based sones. (Figure This 3) isat duethe interfacial to higher CTEsite ( Equationmismatch (4 between)) were found the silica to follow-based a glassessimilar and the titanium substrate, especially for the case of SRT3 (15 mol % incorporated TiO2). Residual trend as the residual strains, with borate-based glasses exhibiting compressive residual stresses and stress inand the the glass titanium coatings substrat (Figuree, especially3) at the for interfacial the case of site SRT3 (Equation (15 mol % (4)) incorporated were found TiO to2). follow Residual a similar stresssilica- basedin the glassglasses coating exhibitings (Figure tensile 3) at residual the interfacial stresses site. (Equation (4)) were found to follow a similar trend astrend the residualas the residual strains, strains, with with borate-based borate-based glassesglasses exhibiting exhibiting compressive compressive residual residual stresses stresses and and silica-basedsilica- glassesbased glasses exhibiting exhibiting tensile tensile residual residual stresses. stresses.

Figure 2. Residual strain at the substrate/coating interface using the SRT and BRT glasses as coatings. Scatter bars indicate one standard deviation from the mean. Figure 2. Residual strain at the substrate/coating interface using the SRT and BRT glasses as coatings. Figure 2. Residual strain at the substrate/coating interface using the SRT and BRT glasses as coatings. Scatter bars indicate one standard deviation from the mean. Scatter bars indicate one standard deviation from the mean.

J. Funct. Biomater. 2016, 7, 32 4 of 14 J. Funct. Biomater. 2016, 7, 32 4 of 14 J. Funct. Biomater. 2016, 7, 32 4 of 14

FigureFigureFigure 3 3.. Residual 3. Residual stresses stresses experienced experienced in in the the glass glass coatingcoating at at the the coatin coating/substratecoating/substrateg/substrate interface interface using using the the SRT andSRT BRTand BRT glassesglasses glasses asas coatings.coating as coatings. Scatters. Scatter bars bars indicateindicate indicate oneone standard standard deviation deviationdeviation from fromfrom the the themean. mean.mean.

2.3. Vickers2.3. Vickers Hardness Hardness 2.3. Vickers Hardness A substitution of SiO2 for B2O3 resulted in an increase in the Vickers hardness of the glass, as A substitution of SiO2 for B2O3 resulted in an increase in the Vickers hardness of the glass, as Ashown substitution in Figure of4, statistically SiO2 for B significant2O3 resulted at 5 and in an15 mol increase % of incorporated in the Vickers TiO2 hardness. The incorporation of the glass, shown in Figure 4, statistically significant at 5 and 15 mol % of incorporated TiO2. The incorporation as shownof TiO in2,Figure however,4, statistically did not significantly significant affect at 5 the and hardness 15 mol for % of the incorporated BRT glass series; TiO for2. The the SRT incorporation glass of TiO2, however, did not significantly affect the hardness for the BRT glass series; for the SRT glass of TiOseries,2, however, the addition did not of significantlyTiO2 at 5 mol affect % decreased the hardness the hardness, for the BRT but glass further series; addition for the did SRT not glass series, the addition of TiO2 at 5 mol % decreased the hardness, but further addition did not series,significantly the addition decrease of TiO 2itat. 5 mol % decreased the hardness, but further addition did not significantly significantly decrease it. decrease it.

Figure 4. Vickers hardness for the SRT and BRT glasses. Scatter bars indicate one standard deviation from the mean. Stars and bars show statistical significance (p < 0.05).

FigureF2.4igure. Bulk 4.4. ModeVickerVickers I sCritical hardness Strain for Energy the SRT Release and BRT Rate glasses.glasses. Using Vickers Scatter Indentation bars indicate one standard deviation from the mean. Stars and bars show statistical signiﬁcancesignificance (p < 0.05). The Vickers indentation test was performed to measure the bulk mode I critical strain energy release rate (Equation (6)), and the results are shown in Figure 5. A sample of the SEM image for SRT0 2.4.2.4. Bulk Mode I Critical Strain Energy Release Rate Using Vickers Indentation showing the indent and the cracks emanating from it is presented in Figure 6. Figure 5 also presents Theas reference Vickers points indentation the data test obtained was performedfrom the literature to measure for the the mode bulk I critical mode strain I critical energy strain release energy releaserate rate of (Equationfused(Equation silica (6)),(6)based), and glass the (99.995% results areSiO shown2) and Pyrex inin Figure (heat5 5resistant.. AA samplesample borosilicate ofof thethe SEMSEM glass). imageimage Based forfor on SRT0SRT0 showingprevious the indent studies, and the thefracture cracks toughness emanating KIC and from the it modulus is presented of elasticity in Figure Ec of6 6fused.. FigureFigure silica5 5also glassalso presents presentsand Pyrex are 0.80 MPa·m1/2 [40] and 0.63 MPa·m1/2 [41], and 72 GPa [40] and 67 GPa [42], respectively. as reference points the data obtained from the literature for the mode I critical strain energy release rate of fused silica basedbased glassglass (99.995%(99.995% SiO22) and Pyrex (heat resistant borosilicate glass). Based on previous studies, thethe fracturefracture toughnesstoughness KKICIC and the modulus of elasticity Ecc of fused silica glass and Pyrex are 0.800.80 MPaMPa··mm1/21/2 [40][40] and and 0.63 0.63 MPa MPa··mm1/21/2 [41][41, ],and and 72 72 GPa GPa [40] [40 ]and and 67 67 GPa GPa [42] [42],, respectively. respectively.

J. Funct. Biomater. 2016, 7, 32 5 of 14 J.J. Funct.Funct. Biomater.Biomater. 2016,, 7,, 3232 5 of 14

IC c The following equation [43],, valid for the plane stress condition, was used to convert these KIC and Ec The following equation [43], valid for the plane2 stress condition, was used2 to convert these KIC and Ec values to the GIC in Figure 5, yielding 8.9 J/m2 for fused silica and 5.9 J/m2 for Pyrex: values to the GIC in Figure 5, yielding 8.9 J/m 2for fused silica and 5.9 J/m for2 Pyrex: values to the GIC in Figure5, yielding 8.9 J/m for fused silica and 5.9 J/m for Pyrex: K G = K (1) G = 2 (1) KE2IC = EICIC GICIC (1) IC Ec c

FFigureigure 5.5. BBulkulk Mode I critical strain energy release rates forfor thethe SRTSRT andand BRTBRT glasses.glasses. TheThe GGIC values Figure 5. Bulk Mode I critical strain energy release rates for the SRT and BRT glasses. The GICIC values forfor fusedfused silicasilica glass glass and and Pyrex Pyrex obtained obtained from from the the literature literatureliterature [40,41 [40,41]] are also are shown also shown for reference. for reference Scatter.. bars indicate one standard deviation from the mean. Stars and bars show statistical signiﬁcance Scatter bars indicate one standard deviation from the mean.. Stars and bars show statistical (p < 0.05). significance (p < 0.05).

Figure 6 6... SEMSEM of of a a Vickers Vickers indent on SRT0 with the emanating cracks. The average half diameter and crack length are 54.8 μmµm and 187.9 µμm,m, respectively.respectively.

2.5. Coating/ Coating/SubstrateSubstrate System Mode I Critical Strain Energy Release Rate The m modeode I criticalcritical strainstrain energyenergy releaserelease ratesrates forfor thethe coating/substratecoating/substrate system system for for both glass series are shown in Figure 77... SystemsSystems mademade withwith borate-basedborate-based glassesglasses exhibitedexhibited higherhigher criticalcritical strainstrain energy release rates in mode I opposed to the silica-basedsilica-based coatings coatings,,, withwith thethe exception exception ofof SRT1SRT1 and and BRT1 (5 mol % incorporated TiO2), which were statistically equivalent (p < 0.05). As a function of the BRT1 (5 (5 mol mol % % inco incorporatedrporated TiO TiO2),2), which which were were statistically statistically equivalent equivalent (p (

J. Funct. Biomater. 2016, 7, 32 6 of 14

J. Funct. Biomater. 2016, 7, 32 6 of 14 incorporated TiO2. Similarly, for systems made with the borate-based series, there is no signiﬁcant increase (p < 0.05) in the critical strain energy release rate between 0 and 5 mol % incorporated TiO2, increase (p < 0.05) in the critical strain energy release rate between 0 and 5 mol % incorporated TiO2, whereas a significant increase is found between 0 and 15 mol % incorporated TiO2, and between 5 whereas a signiﬁcant increase is found between 0 and 15 mol % incorporated TiO2, and between 5 and and 15 mol % incorporated TiO2. 15 mol % incorporated TiO2.

FigureFigure 7. 7Mode. Mode I criticalI critical strain strainenergy energy releaserelease ratesrates for the coating/substrate systems systems with with SRT SRT and and BRTBRT glasses. glasses. Scatter Scatter bars bars indicate indicate one one standard standard deviation deviation from from the mean. the mean. Stars andStars bars and show barsstatistical show signiﬁcancestatistical significance (p < 0.05). (p < 0.05).

3.3. Discussion Discussion TheThe CTE CTE results results confirmedconfirmed that the the bo borate-basedrate-based glasses glasses possessed possessed CTEs CTEs comparable comparable to that to of that ofTi6Al4V, Ti6Al4V, evidenced evidenced by by the the reduced reduced percentage percentage difference difference between between the the borate borate-based-based glasses glasses and and Ti6Al4V,Ti6Al4V, compared compared to to the the silica-based silica-based glasses.glasses. However, However, the the CTE CTE for for the the proposed proposed silica silica-based-based glassesglasses was was significantly significantlylower lower thanthan whatwhat has been reported reported for for other other similar similar glasses glasses (e.g. (e.g.,, CTE CTE of of Bioglass® 45S5 has been reported to be 15.1 × 10−6/°C over the range from 200 to 400 °C [32]), indicating Bioglass® 45S5 has been reported to be 15.1 × 10−6/◦C over the range from 200 to 400 ◦C[32]), that the proposed silicate formulations would provide better adhesion to the metallic substrate indicating that the proposed silicate formulations would provide better adhesion to the metallic compared to other silica-based glasses. Since the addition of TiO2 increased the CTE, the control substrate compared to other silica-based glasses. Since the addition of TiO increased the CTE, silicate formulation should be analyzed to understand the causes behind the reduced2 CTE. Compared the control silicate formulation should be analyzed to understand the causes behind the reduced CTE. to 45S5, SRT0 contains a higher molar percentage of SiO2 and contains ZnO at 16 mol %, whereas Compared to 45S5, SRT0 contains a higher molar percentage of SiO and contains ZnO at 16 mol %, 45S5 does not include ZnO. Higher SiO2 has been determined to decrease2 the CTE of silica-based whereas 45S5 does not include ZnO. Higher SiO has been determined to decrease the CTE of silica-based glasses [27], which additionally explains how 2substituting SiO2 for TiO2 in the SRT glasses translated glassesinto an [27 increase], which in additionally the CTE. Furthermore, explains how ZnO substituting increase (or inclusion, SiO2 for TiO in this2 in case) the SRT has glassesbeen proven translated to intodecrease an increase the CTE in theof silica CTE.-based Furthermore, glasses [44] ZnO, explaining increase (orthe inclusion,reduced CTE in thisof SRT0 case) compared has been to proven 45S5. to decreaseSince the the CTE CTEs of silica-based of the borate glasses-based [ 44glass], explaininges are lower the than reduced that of CTE the ofsubstrate, SRT0 compared the difference to 45S5. in shrinkageSince the caused CTEs the of glass the borate-based coating to experience glasses arecompressive lower than residual that of stresses the substrate,. Compressive the difference residual in shrinkagestresses are caused beneficial the glass, acting coating to prevent to experience cracks from compressive propagating, residual thus stresses. requiring Compressive higher stress residuales to stressescause coating are beneficial, failure [45 acting–47]. toOn prevent the other cracks hand, fromthe CTE propagating, of the silica thus-based requiring glasses induced higher positive stresses to causeresidual coating stresses, failure promoting [45–47]. Onthe growth the other of hand,cracks theunder CTE loading of the [45,48]. silica-based glasses induced positive residualAs stresses, shown in promoting Figure 5, the the measured growth of bulk cracks GIC underof the control loading glass [45, 48of ].the silica-based series, SRT0, 2 wasAs found shown to inbe Figurecomparable5, the measuredto that of fused bulk silica G IC of glass. the controlHowever, glass as ofthe the percentage silica-based of TiO series, in the SRT0, silica-based glass series was increased, the mean GIC value increased, with significant differences (p was found to be comparable to that of fused silica glass. However, as the percentage of TiO2 in the < 0.05) observed between the SRT0 and SRT3 (15 mol % incorporated TiO2). There is limited literature silica-based glass series was increased, the mean GIC value increased, with significant differences available on the mode I critical strain energy release rate (or fracture toughness) of bioactive glasses, (p < 0.05) observed between the SRT0 and SRT3 (15 mol % incorporated TiO2). There is limited specifically on the effect of the inclusion of TiO2 at the expense of the backbone component. The literature available on the mode I critical strain energy release rate (or fracture toughness) of bioactive authors hypothesize that the observed increase in the bulk GIC as the amount of TiO2 increased can glasses, specifically on the effect of the inclusion of TiO2 at the expense of the backbone component. be attributed to the presence of Ti4+ ions. Such ions have been shown to strengthen glass systems and The authors hypothesize that the observed increase in the bulk G as the amount of TiO increased can to improve their mechanical properties [49], due to their small ionicIC radius and high electrical2 charge be attributed to the presence of Ti4+ ions. Such ions have been shown to strengthen glass systems and to

J. Funct. Biomater. 2016, 7, 32 7 of 14 improve their mechanical properties [49], due to their small ionic radius and high electrical charge [50] that tends to strengthen the bonds in the glass. The bulk GIC of the control glass in the borate series, BRT0, was found not to be significantly different (p < 0.05) from that of SRT0; however, the bulk GIC of the remaining borate-based glasses were lower than their counterpart in the silica-based series. This is not surprising, as previous studies on boro-silicate glasses have shown that the incorporation of B2O3 at the expense of SiO2 can decrease the fracture toughness (directly related to GIC through Equation (1)) [51,52]. Furthermore, similar to the silica-based series, the incorporation of TiO2 in the borate series glasses increased the fracture toughness. In terms of the GIC of the coating/substrate system, the presently used DCB testing method was first published by Matinmanesh et al. [53], establishing the GIC for SRT0-based systems at 6.20 ± 0.60 J/m2. The current work expanded on these findings, determining that an increase in 2 TiO2 content resulted in an increase in critical strain energy release rate, peaking at 12.08 ± 1.72 J/m for the silica-based coatings, and 18.50 ± 1.60 J/m2 for the borate-based coatings. This work supports the previous findings that titanium in the glass coating enhances chemical bonding to the titanium substrate [17], resulting in a larger measured GIC for the SRT3 and BRT3 systems, both with 15 mol % of TiO2 incorporated. Additionally, the trends of the GIC for the bulk glass and the coating/substrate system are similar, i.e., an increase in incorporated TiO2 translated into an increase in GIC. This is similar to the work of Li et al. [54], who studied the effect of the incorporation of strontium oxide into borate-based glass coatings applied to Ti6Al4V substrates, and found that GIC increased as the amount of strontium oxide increased. The residual stress in the SRT0 coating using the measured values of the CTE was estimated to be 9.6 ± 2.4 MPa (Section 2.2). This is consistent (no significant difference, p < 0.05) with the 8.6 ± 1.0 MPa value found by Matinmanesh et al. [53] for the same SRT0 coating/substrate system using measurements of the radius of curvature of the coated assembly. Comparing the presented GIC values in Figure5 with those in Figure7 revealed that the glasses in the silica-based series have higher critical strain energy release rates in bulk form compared to when they are applied to the coating/substrate. On the contrary, in the borate based series, the GIC was higher for the coating/substrate system. This effect may be attributed to the nature of the residual stresses that are created during the coating process, i.e., tensile in silica-based systems and compressive in borate-based systems (Section 2.2), with compressive residual stresses providing additional resistance to crack growth. The bulk hardness of the silica-based glass series decreased as the amount of TiO2 incorporated into the glasses increased, while, in statistical terms, the hardness of the borate-based glasses did not significantly change (p < 0.05) with the addition of TiO2. The hardness of the borate-based glasses however, was significantly higher than the silica-based equivalent glasses at 5 and 15 mol % incorporated TiO2. The decrease in hardness with increasing incorporation of TiO2 is consistent with the observed increase in fracture toughness, which is inversely proportionally to the hardness of the glass (Equation (6)).

4. Materials and Methods

4.1. Glass Preparation Silica-based and borate-based glasses in this study were synthesized (compositions and nomenclature are reported in Table1) and characterized by X-ray diffraction (XRD), differential scanning calorimetry (DSC), and Fourier transform infrared spectroscopy (FTIR), among other techniques [55]. TiO2 was added at the expense of SiO2 for the SRT series and at the expense of B2O3 for the BRT series. The glasses were prepared by weighing out appropriate amounts of analytical grade reagents (Fisher Scientiﬁc, Ottawa, ON, Canada & Sigma-Aldrich, Oakville, ON, Canada), ﬁring in silica crucibles (1400–1500 ◦C for 1 h for the silica-based glasses, 1200 ◦C for 1 h for borate-based glasses), and shock quenching in water. The resulting frit was then ball-milled, and sieved to retrieve glass particulates ≤20 µm. J. Funct. Biomater. 2016, 7, 32 8 of 14

Table 1. Glass formulations (mol %).

Silica-Based Glass Borate-Based Glasses Reagent SRT0 SRT1 SRT3 BRT0 BRT1 BRT3

SiO2 52 47 37 0 0 0 B2O3 0 0 0 52 47 37 CaO 12 12 12 12 12 12 P2O5 6 6 6 6 6 6 Na2O 14 14 14 14 14 14 ZnO 16 16 16 16 16 16 TiO2 0 5 15 0 5 15

4.1.1. Discs Preparation Approximately 200 mg of each glass were pressed into a cylindrical mold with the diameter of 6 mm using a hydraulic press with pressure ranging between 2500 and 3000 psi. The sample thicknesses were 2.84 ± 0.19 mm. The pressed discs were then heat treated to promote the coalescence of glass particles and create a sturdy solid to be used for CTE and hardness testing. The heat treatment consisted of ﬁring the discs at temperature Tcoat (Table2) for 15 min, then allowing them to cool down to room temperature.

4.1.2. Coating Preparation Coatings were prepared following the procedure developed by Matinmanesh et al. [53] for enameling Ti6Al4V with bioactive glasses in an ethanol-based suspension. Ti6Al4V substrate samples were degreased and cleaned in ethanol prior to coating. For each glass formulation, a suspension of the glass powder in ethanol (ratio of 5:1, ethanol to glass mass) was deposited on the substrates. Afterwards, the coatings were allowed to air-dry for 30 min, and were then ﬁred at temperature Tcoat, ranging between the glass transition temperature (Tg) and the crystallization temperature (Tx) of each glass (Table2), for 15 min.

Table 2. Glass transition, crystallization, and coating temperatures.

◦ ◦ ◦ Glass Tg ( C) Tx ( C) Tcoat ( C) SRT0 619 735 650 SRT1 592 670 630 SRT3 610 705 640 BRT0 521 603 520 BRT1 530 625 550 BRT3 523 633 550

4.2. Coefﬁcient of Thermal Expansion (CTE) Measurement by Linear Dilatometry The CTE of each glass was tested based on the current ASTM E228 “Standard Test Method for Linear Thermal Expansion of Solid Materials with a Push-Rod Dilatometer” [56]. Samples were prepared following the procedure outlined in Section 4.1.1, with samples measuring 6 mm in diameter and 12 mm in height (by stacking 4 discs), and tested with a Netzsch DIL 402 PC dilatometer (Netzsch Instruments, Burlington, MA, USA). A heating rate of 4 ◦C/min was employed, with a testing temperature range from 25 to 300 ◦C for both the glass series. Based on the measured lengths and temperature changes, CTE was determined as

1 ∆L [αm]Ti = (2) ∆T L0 Ti J. Funct. Biomater. 2016, 7, 32 9 of 14

where αm is the mean CTE of the glass, ∆T is the change in temperature with respect to the initial temperature, L0 is the initial length of the test specimen, and ∆L is the change in length of the sample with respect to the initial length L0.

4.3. Residual Stress and Strain Analysis Due to the mismatch between the CTE of the substrate and of the coating, residual stresses at the interface were induced. Based on the measured CTE as per Section 4.2, and based on the cooling profile and temperatures as per Section 4.1.2, residual strains were computed as εres = αglass − αTi6Al4V (Tcoat − Ti) (3) where εres is the calculated residual strain, αglass is the CTE of the glass, αTi6Al4V is the CTE of the −6 ◦ titanium substrate (9.5 × 10 / C), Tcoat is the coating temperature (from Table2), and T i is the room temperature (25 ◦C). This approach has already been proposed and verified by Oel and Frechette [57]. To determine the residual stresses, Yu et al. [58] proposed the use of beam theory on bi-layer materials with different CTEs subjected to thermal loading. The residual stress at the interface experienced by the glass coating is tc P(ts + tc) 2 σres_glass = (4) 2I + 2EsIs c Ec where εres P = 2 (5) 1 + 1 + (ts+tc) EcAc EsAs 4(EcIc+EsIs) and ts is the thickness of the titanium substrate (3.15 mm), tc is the thickness of the glass coating 3 w ts (90 µm), Is is the second moment of area of the titanium substrate, defined as Is = 12 where w is the width of the titanium substrate and of the glass coating (11 mm), Ic is the second moment of area of the 3 w tc glass coating, defined as Ic = 12 Es is the modulus of elasticity of the titanium substrate (110 GPa), Ec is the modulus of elasticity of the glass coating (35 GPa [59]), As is the cross-sectional area of the titanium substrate, defined as As = w ts, and Ac is the cross-sectional area of the glass coating, defines as Ac = w tc.

4.4. Vickers Hardness Samples (n = 3) for hardness testing were prepared as described in Section 4.1.1. An HM-114 Mitutoyo Testing Machine (Mitutoyo, Mississauga, ON, Canada) was utilized, equipped with a Vickers indenter, to load the samples with a force of 1 kgf (9.81 N) for 10 s. The indent diagonals were measured through the integrated optical microscope at 20×.

4.5. Bulk Mode I Critical Strain Energy Release Rate Using Vickers Indentation The mode I critical strain energy release rate of the bulk glasses was measured by indenting glass discs (prepared as per Section 4.1.1 and indented similarly to the process described in Section 4.4). A schematic depiction of the indentation is shown in Figure8. According to Anstis et al. [ 34], the indentation load needs to be large enough to create an indent pattern that is well-deﬁned and cracks that are longer than the indent diameter (2a), yet shorter than one tenth of the thickness of the sample (300 µm in this case) to avoid interactions with the lower free surface of the specimen. The indentation load is considered too large if it breaks the sample or causes a chipping on the sample’s surface [34]. J. Funct. Biomater. 2016, 7, 32 10 of 14 J. Funct. Biomater. 2016, 7, 32 10 of 14

FFigureigure 8 8.. SchematicSchematic depiction depiction of the cracks cracks emanating emanating from a Vickers indent. a is half of the diameter lengthlength of the dent, and c is the crack length measured from thethe centercenter ofof thethe indent.indent.

For eacheach glass glass composition, composition three, three discs discs were were made. made. By trial By and trial error, and the appropriateerror, the appropriate indentation indentationload to meet load the aforementionedto meet the aforementioned force criteria wasforce found criteria to bewas 3 kgf found (29.43 to N).be This3 kgf indentation (29.43 N). This load indentationwas applied load using was a Macro applied indenter using a HM-114 Macro indenter Mitutoyo HM Testing-114 Mitutoyo Machine (Mitutoyo,Testing Machine Mississauga, (Mitutoyo, ON, Mississauga,Canada) normal ON, to Canada) the surface normal of the to glass the surface for a duration of the glass of 10 for s. The a duration length of of the 10 cracks. The was length measured of the crackthrough was scanning measured electron through microscope scanning (SEM)electron using microscope a JEOL JSM-6380LV(SEM) using SEMa JEOL (JEOL, JSM Peabody,-6380LV SEM MA, (JEOL,USA). ThePeabody, mode MA, I critical USA). strain The mode energy I critical release strain rate of energy the bulk release glass rate was of found the bulk using glass the was equation found usingbelow the [34 ]:equation below [34]: α2P2 GGIC == (6(6)) H2 c c23 IC α 𝑃𝑃 where P is the applied load, H is hardness of the glass 3 (as measured per Section 4.4),4.4), c is the length of the surface trace of the half penny crack measured from the center of the indent, and α is the calibration constantconstantα ==0.0160.016±±0.0040.004. Equation. Equation (6) is(6 derived) is derived for plain for plain stress. stress. Even thoughEven though the current the currentapplication application does not does completely not completely satisfy thesatisfy plain the stress plain condition, stress condition, Equation Eq (6)uation can still(6) can be usedstill be as 𝛼𝛼 usedan approximation as an approximation of GIC given of GIC that given the that thickness the thickness of the discsof the is discs less is than less half than of half their of diameter.their diameter. 4.6. Coating/Substrate System Mode I Critical Strain Energy Release Rate 4.6. Coating/Substrate System Mode I Critical Strain Energy Release Rate The mode I critical strain energy release rate (GIC) of the coating on the substrate was evaluated The mode I critical strain energy release rate (GIC) of the coating on the substrate was evaluated (n = 5) following the procedure outlined by Matinmanesh et al. [53]. Typical sample dimensions for (n = 5) following the procedure outlined by Matinmanesh et al. [53]. Typical sample dimensions for the bi-layer double-cantilever beam (DCB) specimen are shown in Figure9. Three samples per glass the bi-layer double-cantilever beam (DCB) specimen are shown in Figure 9. Three samples per glass composition were tested. Coated samples (prepared as described in Section 4.1.2) were used to make composition were tested. Coated samples (prepared as described in Section 4.1.2) were used to make the test specimens, then an epoxy layer (J-B Weld 8265-S Cold Weld Compound, Sulphur Springs, TX, the test specimens, then an epoxy layer (J-B Weld 8265-S Cold Weld Compound, Sulphur Springs, USA) was deposited to cover the glass and attach the second titanium alloy substrate. Specimens were TX, USA) was deposited to cover the glass and attach the second titanium alloy substrate. Specimens loaded using a STM United Tensile Tester (United Testing Systems, Inc., Huntington Beach, CA, USA) were loaded using a STM United Tensile Tester (United Testing Systems, Inc., Huntington Beach, using a 500-N load cell at a rate of 0.5 mm/min; then based on the recorded loads, the mode I critical CA, USA) using a 500-N load cell at a rate of 0.5 mm/min; then based on the recorded loads, the mode strain energy release rate GIC was calculated as: I critical strain energy release rate GIC was calculated as: 1212LL2a2 GG = (7) IC 22 23 (7) EEsww tts IC 2 3 where L refers to the load to start the crack, a is thes cracks length, Es to the tensile modulus of the where L refers to the load to start the crack, a is the crack length, Es to the tensile modulus of the substrate (110 GPa), w to the specimen width (11 mm), and ts is the thickness of the substrate (3.15 substrate (110 GPa), w to the specimen width (11 mm), and ts is the thickness of the substrate (3.15 mm). mm).

J.J. Funct. Funct. Biomater. Biomater. 20162016,, 77,, 32 32 1111 of of 14 14

FFigureigure 9 9.. BiBi-layer-layer double double cantilever beam specimens specimens.. All All units units are in millimeters.millimeters. Gray Gray materials representrepresent the the titanium titanium alloy alloy substrates, substrates, black black material material represents represents the the glass glass,, and and white white material material representsrepresents the the epoxy epoxy..

4.74.7.. Statistical Statistical Methods Methods TheThe resultsresults ofof all all the the measurements measurements were were expressed expressed as means as withmeans experimental with experimental scatter expressed scatter expressedas a standard as a deviation. standard Additionally,deviation. Additionally, one-way analysis one-way of varianceanalysis (ANOVA)of variance was (ANOVA) employed was to employedanalyze the to dataanalyze to determine the data to the determine signiﬁcance the significance in mean difference in mean across difference the gatheredacross the data gathered when datap < 0.05 when. Post-hoc p < 0.05. Tukey Post test-hoc was Tukey used test on MiniTabwas used 17 on (MiniTab MiniTab Inc., 17 State(MiniTab College, Inc., PA, State USA). College, The Tukey PA, USAtest assumes). The Tukey equal test variance assumes in theequal data variance sets being in analyzedthe data sets to determine being analyzed the signiﬁcance to determine in mean the significancedifference across in mean all factorsdifference (i.e., across all glasses all factors in both (i.e. series)., all glasses in both series).

5.5. Conclusions Conclusions SilicaSilica-based-based and and borate borate-based-based glasses glasses were were synthesized synthesized and and characterized characterized in in terms terms of of their their mechanicalmechanical properties properties relevant relevant to totheir their use use as metallic as metallic coating coating materials. materials. It was Itobserved was observed that borate that- basedborate-based glasses glassesexhibited exhibited CTE values CTE valuesthat were that closer were to closer the sub to thestrate’s substrate’s (Ti6Al4V) (Ti6Al4V) CTE, a CTE, common a common alloy usedalloy in used medical in medical implants; implants; this translated this translated into higher into highermode I modecritical I criticalstrain energy strain energyrelease releaserates for rates the boratefor the-based borate-based glasses and glasses compressive and compressive residual residualstresses and stresses strains and at strains the coating/substrate at the coating/substrate interface, interface, outperforming the silica-based glasses counterpart. An increase in the content of TiO in outperforming the silica-based glasses counterpart. An increase in the content of TiO2 in the glasses2 resultedthe glasses in an resulted increase in in an the increase mode inI critical the mode strain I critical energy strain release energy rate for release both the rate bulk for bothglass the and bulk for glass and for the coating/substrate system. Borate-based glass BRT3, with 15 mol % TiO incorporated, the coating/substrate system. Borate-based glass BRT3, with 15 mol % TiO2 incorporated,2 exhibited superiorexhibited properties superior propertiesoverall compared overall comparedto the other to proposed the other glasses proposed in glassesthis work, in thisas well work, as asBioglass well as® ® 45S5Bioglass and Pyrex.45S5 and Pyrex. Acknowledgments: The authors would like to thank the Collaborative Health Research Project fund Acknowledgments:(#315694-DAN) for financing The authors this would research like and to the thank Ryerson the Collaborative University Strategic Health HireResearch program Project for earlyfund assistance(#315694- DAN)with Rodriguez’s for financing stipend. this research The authors and the would Ryerson also likeUniversity to thank Strategic Isaac Beniluz Hire andprogram Bharath for Krishnanearly assistance for assisting with Rodriguez’swith preparing stipend. samples The authors for the would CTE and also coating like to thank GIC tests.Isaac Beniluz Finally, and Canada Bharath Research Krishnan Chairs for assisting is gratefully with preparingacknowledged samples for the for salary the supportCTE and of Marcellocoating Papini.GIC tests. Finally, Canada Research Chairs is gratefully acknowledgedAuthor Contributions: for the salaryOmar support Rodriguez, of Marcello Ali Matinmanesh, Papini. Emil H. Schemitsch, Paul Zalzal, Marcello Papini, and Mark R. Towler conceived and designed the experiments; Omar Rodriguez, Ali Matinmanesh, Sunjeev Phull, Authorand Owen Contributions: Clarkin performed Omar Rodriguez, the experiments; Ali Matinmanesh, Omar Rodriguez Emil andH. Schemitsch, Ali Matinmanesh Paul Zalzal, analyzed Marcello the data Papini and, andwrote Mark the paper.R. Towler conceived and designed the experiments; Omar Rodriguez, Ali Matinmanesh, Sunjeev Phull, and Owen Clarkin performed the experiments; Omar Rodriguez and Ali Matinmanesh analyzed the data Conflicts of Interest: The authors declare no conflict of interest. The founding sponsors had no role in the design andof the wrote study; the inpaper. the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the Conflictsdecision to of publish Interest: the The results. authors declare no conflict of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decisionReferences to publish the results. 1. Webster, J.B.; Chou, T.; Kenly, M.; English, M.; Roberts, T.L.; Bloebaum, R.D. Perceptions and acceptance Referencesof osseointegration among individuals with lower limb amputations: A prospective survey study. 1. Webster,J. Prosthet. J.B. Orthot.; Chou,2009 T.;, Kenly,21, 215–222. M.; English, [CrossRef M.;] Roberts, T.L.; Bloebaum, R.D. Perceptions and acceptance 2. ofSullivan, osseointegration J.; Uden, M.; among Robinson, individuals K.P.; Sooriakumaran, with lower limb S. Rehabilitation amputations of: A the prospective trans–femoral survey amputee study. with J. Prosthetan osseointegrated. Orthot. 2009 prosthesis:, 21, 215–222. The United Kingdom experience. Prosthet. Orthot. Int. 2003, 27, 114–120. 2. Sullivan,[CrossRef J.;][ Uden,PubMed M.;] Robinson, K.P.; Sooriakumaran, S. Rehabilitation of the trans–femoral amputee with 3. anTillander, osseointegrated J.; Hagberg, prosthesis: K.; Hagberg, The United L.; Branemark, Kingdom R. experience. Osseointegrated Prosthet titanium. Orthot implants. Int. 2003 for, 27 limb, 114 prostheses–120. Clin. Orthop. Relat. Res. 468 3. Tillander,attachments: J.; InfectiousHagberg, complications.K.; Hagberg, L.; Branemark, R. Osseointegrated2010, , 2781–2788. titanium [CrossRef implants][ PubMedfor limb] prostheses attachments: Infectious complications. Clin. Orthop. Relat. Res. 2010, 468, 2781–2788.

J. Funct. Biomater. 2016, 7, 32 12 of 14

4. Pendegrass, C.J.; Goodship, A.E.; Blunn, G.W. Development of a soft tissue seal around bone-anchored transcutaneous amputation prostheses. Biomaterials 2006, 27, 4183–4191. [CrossRef][PubMed] 5. Chaudhari, A.; Braem, A.; Vleugels, J.; Martens, J.A.; Naert, I.; Cardoso, M.V.; Duyck, J. Bone tissue response to porous and functionalized titanium and silica based coatings. PLoS ONE 2011, 6, e24186. [CrossRef] [PubMed] 6. Buser, D.; Schenk, R.K.; Steinemann, S.; Fiorellini, J.P.; Fox, C.H.; Stich, H. Inﬂuence of surface characteristics on bone integration of titanium implants. A histomorphometric study in miniature pigs. J. Biomed. Mater. Res. 1991, 25, 889–902. [CrossRef][PubMed] 7. Hench, L.L.; Splinter, R.J.; Allen, W.C.; Greenlee, T.K. Bonding mechanisms at the interface of ceramic prosthetic materials. J. Biomed. Mater. Res. 1971, 5, 117–141. [CrossRef] 8. Chen, M.; Li, W.; Shen, M.; Zhu, S.; Wang, F. Glass coatings on stainless steels for high-temperature oxidation protection: Mechanisms. Corros. Sci. 2014, 82, 316–327. [CrossRef] 9. Sola, A.; Bellucci, D.; Cannillo, V. Enamelled coatings produced with low-alkaline bioactive glasses. Surf. Coat. Technol. 2014, 248, 1–8. [CrossRef] 10. Saiz, E.; Goldman, M.; Gomez-Vega, J.M.; Tomsia, A.P.; Marshall, G.W.; Marshall, S.J. In vitro behavior of silicate glass coatings on Ti6Al4V. Biomaterials 2002, 23, 3749–3756. [CrossRef] 11. Peddi, L.; Brow, R.K.; Brown, R.F. Bioactive borate glass coatings for titanium alloys. J. Mater. Sci. Mater. Med. 2008, 19, 3145–3152. [CrossRef][PubMed] 12. Flaten, T.P.; Alfrey, A.C.; Birchall, J.D.; Savory, J.; Yokel, R.A. Status and future concerns of clinical and environmental aluminum toxicology. J. Toxicol. Environ. Health 1996, 48, 527–542. [PubMed] 13. Boyd, D.; Li, H.; Tanner, D.A.; Towler, M.R.; Wall, J.G. The antibacterial effects of zinc ion migration from zinc-based glass polyalkenoate cements. J. Mater. Sci. Mater. Med. 2006, 17, 489–494. [CrossRef][PubMed] 14. Coughlan, A.; Scanlon, K.; Mahon, B.P.; Towler, M.R. Zinc and silver glass polyalkenoate cements: An evaluation of their antibacterial nature. Biomed. Mater. Eng. 2010, 20, 99–106. [PubMed] 15. Murphy, S.; Wren, A.W.; Towler, M.R.; Boyd, D. The effect of ionic dissolution products of Ca–Sr–Na–Zn–Si bioactive glass on in vitro cytocompatibility. J. Mater. Sci. Mater. Med. 2010, 21, 2827–2834. [CrossRef] [PubMed] 16. Foley, J.; Blackwell, A. Ion release from copper phosphate cement and inﬂuence on Streptococcus mutans growth in vitro: A comparative study. Caries Res. 2002, 37, 416–424. [CrossRef] 17. Sola, A.; Bellucci, D.; Cannillo, V.; Cattini, A. Bioactive glass coatings: A review. Surf. Eng. 2011, 27, 560–572. [CrossRef] 18. Welsch, G.; Boyer, R.; Collings, E.W. Materials Properties Handbook: Titanium Alloys; ASM International: Materials Park, OH, USA, 1993; p. 516. 19. Schrooten, J.; Helsen, J.A. Adhesion of bioactive glass coating to Ti6Al4V oral implant. Biomaterials 2000, 21, 1461–1469. [CrossRef] 20. Mehdipour, M.; Afshar, A.; Mohebali, M. Electrophoretic deposition of bioactive glass coating on 316L stainless steel and electrochemical behavior study. Appl. Surf. Sci. 2012, 258, 9832–9839. [CrossRef] 21. Fiorilli, S.; Baino, F.; Cauda, V.; Crepaldi, M.; Vitale-Brovarone, C.; Demarchi, D.; Onida, B. Electrophoretic deposition of mesoporous bioactive glass on glass-ceramic foam scaffolds for bone tissue engineering. J. Mater. Sci. Mater. Med. 2015, 26, 1–12. [CrossRef][PubMed] 22. Ponsot, I.; Detsch, R.; Boccaccini, A.R.; Bernardo, E. Waste derived glass ceramic composites prepared by low temperature sintering/sinter-crystallisation. Adv. Appl. Ceram. 2015, 114 (Suppl. 1), S17–S25. [CrossRef] 23. Brow, R.K.; Watkins, R.D. Reactions and Bonding between Glasses and Titanium; No. SAND-87-0541C; CONF-871207-18; Sandia National Labs.: Albuquerque, NM, USA, 1987.

24. Saha, S.K.; Jain, H.; Goldstein, J.I.; Miller, A.C.; Brow, R.K. Reaction between titanium and B2O3 melt/glass. Phys. Chem. Glasses 1998, 39, 118–121. 25. Branemark, R.; Branemark, P.I.; Rydevik, B.; Myers, R.R. Osseointegration in skeletal reconstruction and rehabilitation: A review. J. Rehabil. Res. Dev. 2001, 38, 175–182. [PubMed] 26. Branemark, P.I.; Hansson, B.O.; Adell, R.; Breine, U.; Lindstrom, J.; Hallen, O.; Ohman, A. Osseointegrated implants in the treatment of the edentulous jaw. Scand. J. Plast. Reconstr. Surg. Suppl. 1997, 11, 1–52. 27. Pavon, J.; Jimenez-Pique, E.; Anglada, M.; Lopez-Esteban, S.; Saiz, E.; Tomsia, A.P. Stress–corrosion cracking by indentation techniques of a glass coating on Ti6Al4V for biomedical applications. J. Eur. Ceram. Soc. 2006, 26, 1159–1169. [CrossRef] J. Funct. Biomater. 2016, 7, 32 13 of 14

28. Lotﬁbakhshaiesh, N.; Brauer, D.S.; Hill, R.G. Bioactive glass engineered coatings for Ti6Al4V alloys: Inﬂuence of strontium substitution for calcium on sintering behaviour. J. Non-Cryst. Solids 2010, 356, 2583–2590. [CrossRef] 29. Charles, L.F.; Shaw, M.T.; Olson, J.R.; Wei, M. Fabrication and mechanical properties of PLLA/PCL/HA composites via a biomimetic, dip coating, and hot compression procedure. J. Mater. Sci. Mater. Med. 2010, 21, 1845–1854. [CrossRef][PubMed] 30. Lardner, T.J.; Ritter, J.E.; Shiao, M.L.; Lin, M.R. Behavior of indentation cracks near free surfaces and interfaces. Int. J. Fract. 1990, 44, 133–143. [CrossRef] 31. Gomez-Vega, J.M.; Saiz, E.; Tomsia, A.P.; Marshall, G.W.; Marshall, S.J. Bioactive glass coatings with hydroxyapatite and Bioglass® particles on Ti-based implants. 1. Processing. Biomaterials 2000, 21, 105–111. [CrossRef] 32. Lopez-Esteban, S.; Saiz, E.; Fujino, S.; Oku, T.; Suganuma, K.; Tomsia, A.P. Bioactive glass coatings for orthopedic metallic implants. J. Eur. Ceram. Soc. 2003, 23, 2921–2930. [CrossRef] 33. Lawn, B.R.; Evans, A.G.; Marshall, D.B. Elastic/plastic indentation damage in ceramics: The median/radial crack system. J. Am. Ceram. Soc. 1980, 63, 574–581. [CrossRef] 34. Anstis, G.R.; Chantikul, P.; Lawn, B.R.; Marshall, D.B. A critical evaluation of indentation techniques for measuring fracture toughness: I, direct crack measurements. J. Am. Ceram. Soc. 1981, 64, 533–538. [CrossRef] 35. Pharr, G.M. Measurement of mechanical properties by ultra-low load indentation. Mater. Sci. Eng. A Struct. Mater. 1998, 253, 151–159. [CrossRef] 36. Fett, T. Computation of the Crack Opening Displacements for Vickers Indentation; Forschungszentrum Karlsruhe (FZKA): Karlsruhe, Germany, 2002. 37. Fett, T.; Burghard, Z.; Zimmermann, A.; Aldinger, F. Residual stresses and stress intensity factors for vickers indentation cracks in glass derived from COD measurements. Adv. Eng. Mater. 2004, 6, 914–918. [CrossRef] 38. Fett, T.; Kounga, A.B.; Rödel, J. Stresses and stress intensity factor from COD of Vickers indentation cracks. J. Mater. Sci. 2004, 39, 2219–2221. [CrossRef] 39. Laugier, M.T. New formula for indentation toughness in ceramics. J. Mater. Sci. Lett. 1987, 6, 355–356. [CrossRef] 40. Wiederhorn, S.M. Fracture surface energy of glass. J. Am. Ceram. Soc. 1969, 52, 99–105. [CrossRef] 41. Harding, D.S.; Oliver, W.C.; Pharr, G.M. Cracking during nanoindentation and its use in the measurement of fracture toughness. MRS Proc. 1994, 356, 663. [CrossRef] 42. Venkateshan, S.P. Mechanical Measurements; John Wiley & Sons: New York, NY, USA, 2015. 43. Anderson, T.L. Fracture Mechanics: Fundamentals and Applications; CRC Press: Boca Raton, FL, USA, 2005.

44. Hurt, J.C.; Phillips, C.J. Structural Role of Zinc Oxide in Glasses in the System Na2O–ZnO–SiO2. J. Am. Ceram. Soc. 1970, 53, 269–273. [CrossRef] 45. Elber, W. Effects of shot-peening residual stresses on the fracture and crack-growth properties of D6AC steel. Fract. Toughness Slow Stable Crack. 1974.[CrossRef] 46. Sglavo, V.M.; Larentis, L.; Green, D.J. Flaw-Insensitive Ion-Exchanged Glass: I, Theoretical Aspects. J. Am. Ceram. Soc. 2001, 84, 1827–1831. [CrossRef] 47. Bermejo, R.; Torres, Y.; Sanchez-Herencia, A.J.; Baudín, C.; Anglada, M.; Llanes, L. Residual stresses, strength and toughness of laminates with different layer thickness ratios. Acta Mater. 2006, 54, 4745–4757. [CrossRef] 48. Hutchinson, J.W.; Suo, Z. Mixed mode cracking in layered materials. Adv. Appl. Mech. 1991, 29, 63–191. 49. Clement, J.; Avila, G.; Navarro, M.; Martinez, S.; Ginebra, M.P.; Planell, J.A. Chemical durability

and mechanical properties of calcium phosphate glasses with the addition of Fe2O3, TiO2 and ZnO. Key Eng. Mater. 2001, 192, 621–624. [CrossRef] 50. Rajendran, V.; Devi, A.G.; Azooz, M.; El-Batal, F.H. Physicochemical studies of phosphate based

P2O5–Na2O–CaO–TiO2 glasses for biomedical applications. J. Non-Cryst. Solids 2007, 353, 77–84. [CrossRef] 51. Yoshida, S.; Tanaka, H.; Hayashi, T.; Matsuoka, J.; Soga, N. Scratch resistance of sodium borosilicate glass. J. Ceram. Soc. Jpn. 2001, 109, 511–515. [CrossRef] 52. Seal, A.K.; Chakraborti, P.; Roy, N.R.; Mukherjee, S.; Mitra, M.K.; Das, G.C. Effect of phase separation on the

fracture toughness of SiO2–B2O3–Na2O glass. Bull. Mater. Sci. 2005, 28, 457–460. [CrossRef] 53. Matinmanesh, A.; Rodriguez, O.; Towler, M.R.; Zalzal, P.; Schemitsch, E.H.; Papini, M. Quantitative evaluation of the adhesion of bioactive glasses onto Ti6Al4V substrates. Mater. Des. 2016, 97, 213–221. [CrossRef] J. Funct. Biomater. 2016, 7, 32 14 of 14

54. Li, Y.; Matinmanesh, A.; Curran, D.J.; Schemitsch, E.H.; Zalzal, P.; Papini, M.; Wren, A.W.; Towler, M.R. Characterization and Fracture Property of Different Strontium-containing Borate-based Glass Coatings for Ti6Al4V Substrates. J. Non-Cryst. Solids 2016. submitted. 55. Rodriguez, O.; Curran, D.J.; Papini, M.; Placek, L.M.; Wren, A.W.; Schemitsch, E.H.; Zalzal, P.; Towler, M.R. Characterization of silica-based and borate-based, titanium-containing bioactive glasses for coating metallic implants. J. Non-Cryst. Solids 2015, 433C, 95–102. [CrossRef] 56. ASTM Standard E228. Standard Test Method for Linear Thermal Expansion of Solid Materials With a Push-Rod Dilatometer; ASTM International: West Conshohocken, PA, USA, 2011. 57. Oel, H.J.; Frechette, V.D. Stress distribution in multiphase systems: I, composites with planar interfaces. J. Am. Ceram. Soc. 1967, 50, 542–549. [CrossRef] 58. Yu, Y.; Ashcroft, I.A.; Swallowe, G. An experimental investigation of residual stresses in an epoxy–steel laminate. Int. J. Adhes. Adhes. 2006, 26, 511–519. [CrossRef] 59. Kokubo, T.; Kim, H.M.; Kawashita, M. Novel bioactive materials with different mechanical properties. Biomaterials 2003, 24, 2161–2175. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).